The cohort of all Australian former army conscripts of the Vietnam conflict ea was followed from 1965 to 1982 to determine mortality rates and causes of death following completion of their National Service. Suiciders were compared with a random sample of survivors using information contained in their military documents in a nested case-control study. Their military document information was recorded before men were selected for Vietnam service and is uncontaminated by "recall bias." Suicide victims had lower mean scores on the army general intelligence and mechanical comprehension tests, were less likely to have continued education beyond high school, were less likely to be employed in white-collar or skilled blue-collar jobs between leaving school and being drafted, and more likely to have volunteered for the draft. They were more likely to have committed a civilian offense before joining the army, more likely to have gone absent without leave (AWOL), and more likely to have committed other offenses during military service. Suiciders were more likely to have a history of diagnosis and treatment for psychological disorder during service and to be judged to be less than emotionally stable at discharge. Service in Vietnam was not associated with suicide. A log-linear regression model was used to analyze death rates associated with five types of variables: cognitive abilities, education, preservice employment, conduct while in service, and physical and mental health. This analysis produced a model containing only four variables: intelligence test score, postschool education, AWOL charge during service, and history of diagnosis and treatment of psychological problems. The difference in death rates between high scorers on these items and low scorers was 46-fold, from 5.2 to 240.9 per 10,000 person-years.